TechCrunch Disrupt 2026 is spotlighting essential discussions on AI safety, featuring five key sessions designed to help founders understand how to build trustworthy AI products. With artificial intelligence increasingly embedded in business workflows, autonomous vehicles, robots, and more, safety and security have become crucial parts of the development process. The event, taking place October 13-15 in San Francisco, brings together leaders from Anthropic, Nvidia, AWS, Waabi, and other top companies to explore the challenges and solutions of moving AI from pilot projects to reliable, real-world deployments.

One of the highlighted sessions, led by Cat de Jong of Anthropic, dives into why some enterprise AI implementations succeed while others remain stuck in pilot mode, offering founders practical insights for transitioning from experimentation to production. Another important discussion addresses agent security with Okta’s Ric Smith and NanoCo’s Gavriel Cohen, focusing on how AI agents' permissions and infrastructure-level security issues must be handled carefully as agentic AI takes on more autonomous roles. AWS and cybersecurity experts will also examine the complexity of securing AI in enterprise cloud environments, a growing hurdle for AI startups pursuing large-scale adoption.

TechCrunch Disrupt also tackles the high-stakes nature of autonomous AI in physical systems through a session featuring Shield AI, General Motors, and Waabi experts. Their conversation centers on testing, validation, regulatory challenges, and cultivating a safety-first mindset for AI deployed in vehicles, aircraft, and industrial applications where failure is not an option. Additionally, Nvidia’s Les Karpas will explore why robotics has yet to experience its own “ChatGPT moment,” spotlighting the data challenges and innovation needed to accelerate physical AI’s capabilities and trustworthiness.

Together, these five AI safety sessions address the critical question for founders: Can people trust what you’re building? With more than 200 sessions and over 10,000 attendees expected, TechCrunch Disrupt 2026 offers a comprehensive platform for AI innovators to engage with safety, security, and reliability topics essential for gaining enterprise and consumer trust. Founders are encouraged to register by September 25 to save up to $200 and take advantage of a 50% discount on a second ticket, securing their spot to learn from top AI leaders and prepare their startups for successful, safe AI adoption.